By Liang Wei, Senior Application Engineer, EVST — collaborative assembly and fastening cells.
Last updated: 16 June 2026.
Answer first: For high-mix, low-volume assembly, a collaborative robot drops the safety fence (after an ISO/TS 15066 risk assessment), changes product in minutes via drag-to-teach and a program library, and tightens every fastener to a closed-loop torque target with a per-bolt record. It fits when you have many products in small batches, frequent changeover, traceable-torque requirements, or a floor too tight for a fenced cell — cases where a dedicated rig is scrapped on every product change.
What is collaborative flexible assembly?
It’s an assembly cell built around a collaborative robot (cobot) that shares the workspace with operators instead of being locked behind a fence. The cobot does the repeatable work — locate, pick, assemble, tighten, confirm — while a person handles the flexible, judgment tasks alongside it. “Flexible” means the same cell runs many products: you change the job by recalling a program, not by rebuilding tooling.
Why dedicated tooling hits a ceiling on high-mix work
A dedicated rig is efficient for one product at volume — and a poor fit for many products in small batches.
First, changeover is expensive. Hard tooling for product A is scrapped or re-set when product B arrives; the engineering and downtime per change eat the economics.
Second, torque rides on feel. Manual fastening leaves no record of the value reached — exactly what quality systems want you to prove.
Third, volume can’t amortize the rig. Low batches never pay back heavy fixed automation.
Fourth, fences eat floor. A fenced industrial cell needs space many high-mix shops don’t have.
How a collaborative cell changes the math
Fence-free (after ISO/TS 15066). Power- and force-limited operation, validated by a task risk assessment, lets the cobot work next to people — saving the floor a fence would consume. Fence-free is an engineered outcome, not a default: change the part, tool or takt and the assessment is redone.
Minute-scale changeover. Drag-to-teach sets a path in minutes; products live in a program library and are recalled in one tap. High mix runs on one line.
Per-bolt torque traceability. Tightening is closed-loop on torque and every value is logged — IATF 16949-compatible evidence as a by-product of running.
Repeatable precision. Typical repeatability around ±0.03 mm supports consistent fastening and placement across the mix.
Manual/dedicated vs collaborative cell
| Dimension | Dedicated rig / manual | Collaborative cell |
|---|---|---|
| Changeover | Re-tool / scrap fixture | Minutes (drag-teach + library) |
| Torque record | None / by feel | Closed-loop, per-bolt log |
| Footprint | Fenced, large | Fence-free (ISO/TS 15066) |
| Low-volume economics | Hard to amortize | Pays on mix + traceability |
| Human role | Replaced or manual | Shared station |
Typical of well-designed cells; validate against your parts and a task risk assessment.
When a collaborative cell pays off
- Many products, small batches — the core case.
- Frequent changeover — minute-scale recall beats re-tooling.
- Torque must be traceable — closed-loop + logging is the reliable path.
- Tight floor — no fence, shared station.
If you run one product at high volume with no traceability need, a dedicated rig may still win. Match the cell to the mix.
Where it fits: cross-industry
Sprocket flanges and hydraulic valve blocks, 3C devices, appliances, and e-drive housings — anywhere a changing part mix needs flexible fastening and assembly with a record. The fixtures change; the cell does not.
Standards and references that frame the design
- ISO/TS 15066 — collaborative-robot safety: force and pressure limits by body region, the basis for fence-free operation.
- ISO 10218-1 / -2 — industrial-robot and cell safety / integration.
- IATF 16949 / ISO 9001 — quality systems that require the per-bolt torque traceability the cell produces.
Pre-deployment checklist
- List the part family and batch sizes; confirm high-mix fit.
- Define the program library and changeover targets.
- Specify closed-loop torque targets and the per-bolt record format.
- Run the ISO/TS 15066 task risk assessment for fence-free operation.
- Confirm shared-station layout and operator interaction.
Frequently asked questions
Is fence-free actually safe? It can be, after an ISO/TS 15066 task risk assessment with power/force limits and collision detection. It is task-specific — change the part, tool or takt and re-assess; never copy-paste.
How fast is changeover? Minute-scale: drag-to-teach plus a program library recall, versus re-tooling a dedicated rig.
Is the torque traceable? Yes — closed-loop on torque with a per-bolt log, compatible with IATF 16949.
Cobot vs a dedicated machine? Cobot for high-mix, low-volume, frequent changeover or tight floors; dedicated machine for single-product high volume.
What precision can it hold? Typically around ±0.03 mm repeatability — validate against your parts.
Key takeaways
- A collaborative cell makes high-mix assembly fence-free, minute-changeover, and torque-traceable.
- It pays off on mix, changeover frequency, traceability needs, or tight floors.
- Fence-free is an ISO/TS 15066 outcome, re-assessed per task — not a default.
Talk to EVST about your cell
Send us your part family, batch sizes and torque requirements — we’ll design the program library, fastening and a fence-free layout assessed to ISO/TS 15066.
→ Contact us for a flexible-assembly assessment.
Or reach us directly: sales@evsrobot.com · Tel / WhatsApp / WeChat: +86 19381626253
Related reading: robotic machine tending, vision error-proofing assembly, and torque traceability.